OpenAI crisis: Altman out?

Two former OpenAI board members–Helen Toner and Tasha McCauley–who were behind the original campaign to push OpenAI CEO, Sam Altman, out in November (and left following his reinstatement)--have published damning statements, calling for greater transparency and Government regulation, claiming Altman's leadership is ruining the company.

What are Toner and McCauley saying about Altman?

Both ex-board members are standing by their original decision to oust Altman, last year, sharing that multiple senior leaders at OpenAI came to the board with “grave concerns” about the “toxic culture of lying” cultivated by Altman, that “can be characterized as psychological abuse.”

They claim that Altman had exhibited  "long-standing patterns of behavior” which "undermined the board’s oversight of key decisions and internal safety protocols”, which, combined with OpenAI’s reliance on self-governance (due to a lack of Government regulation), is a “recipe for Artificial General Intelligence (AGI)”.

They believe that OpenAI can’t hold itself accountable with Altman as CEO, as developments since his reinstatement–including the rapid exit of key safety team members–“bodes ill for the OpenAI experiment in self-governance.”

As well as suggesting that Altman should go, they’ve also called for greater Government intervention and regulatory frameworks for companies, like OpenAI, that develop AI models, stressing that self-governance can’t withstand profit-driven pressures.  

"Even with the best of intentions, without external oversight, this kind of self-regulation will end up unenforceable, especially under the pressure of immense profit incentives.”

Why are they saying this?

Since Altman’s return in November, OpenAI has been hit with several public scandals, including the shock disbandment of its safety team, with ex-team members claiming that Altman prioritized “shiny new products” over safety, leaving them unable to do their job. Altman was also recently called out by actress Scarlett Johansson for stealing her voice, without permission, for the voice feature in its new GPT-4o model, and reports have been circling that Altman forces employees to either sign a highly-restrictive exit document or lose millions of dollars of earned equity when they leave.